Submission + - Why AI token prices are about to plummet (businessinsider.com)

Submitted by ZipNada
ZipNada writes: The main force driving token prices lower is a new wave of technology that's sweeping through AI data centers.

Nvidia's Blackwell GPUs are being installed in huge volumes right now. By the second half of this year, these systems, which are really supercomputers rather than chips, will be operating at scale, helping AI labs train new models and run them more efficiently.

These systems took a while to install properly, partly because they needed to be water-cooled and required other gnarly new data center setups. But the payoff could be huge.

50 x more, 35 x cheaper
SemiAnalysis, a respected AI research firm, compared Nvidia's top Blackwell system, the GB 300 NVL72, to Nvidia's previous system, called the Hopper HGX 200.

With the older system, each GPU generated 90 tokens per second, while the new Blackwell system generated 6,000. That's 65 times more.

Submission + - Anthropic Shutdown Raises Questions About AI Vendor Dependence (nerds.xyz)

Submitted by BrianFagioli
BrianFagioli writes: Following the U.S. governmentâ(TM)s directive that led Anthropic to suspend access to some of its newest AI models, it can be argued that the incident highlights a broader risk facing businesses increasingly dependent on AI services. The article suggests that organizations building critical workflows around a single AI provider may be exposing themselves to regulatory, operational, and business continuity risks. The article also questions whether companies are becoming too reliant on AI itself, rather than treating it as a tool that augments human expertise. What contingency plans should organizations have if a key AI service becomes unavailable overnight?

Submission + - Zuckerberg on Meta's North Star: "The Most Talented People in the World" Matter

Submitted by theodp
theodp writes: In Meta’s New AI Unit Is a Total Mess (alt source), WIRED reports that executives and employees alike are struggling with Meta’s chaotic AI strategy, according to sources and internal discussions it reviewed. On Friday, CEO Mark Zuckerberg in a memo addressed widespread dissatisfaction with how Meta assembled its Applied AI unit of about 6,500 "draftee" engineers and product managers and the drudgework they allege they have been assigned to improve AI models.

Zuckerberg suggested the team was a waypoint, not a destination. “Work like AAI is critical to advancing our models and it lets very talented people contribute to those efforts while we create other roles they can contribute to around Meta over the coming months as well,” he wrote. "Meta’s north star is to be the best place for the most talented people in the world to make an impact." Last month, it was reported that Zuckerberg purportedly explained on leaked audio that Meta settled on tracking its own employees' keystrokes, mouse clicks, and screenshots for AI training, saying: “We are using this to feed a very large amount of content into the AI model, so that way it can learn how smart people use computers to accomplish tasks. I think that this is going to be a very big advantage if we can do it.” He added: "The average intelligence of the people who are at this company is significantly higher than the average set of people."

Zuckerberg also reiterated a vow to not carry out additional mass layoffs this year. He introduced a plan to limit the number of employees per manager, which on some teams, such as Applied AI, had deliberately ballooned to a ratio of 50 to one. Budgets for team events would increase, he said, and a large hackathon planned for next month could also help bring the company together. By the end of the year, employees in many locations would have assigned desks again, the CEO wrote.

Submission + - US Federal Government bars foreigners from using Anthropic's latest AI models (nytimes.com)

Submitted by ormembar
ormembar writes: The New York Times reports that the US Federal Government has barred Anthropic from allowing any foreigners, including inside the company, to use their latest AI models. Since Anthropic cannot technically enforce such request, they have suspend access to Fable 5 and Mythos 5 until further notice. This could have serious implications on any new models released by other AI giants in the US like OpenAI, Google, Meta, Microsoft, etc.

> Anthropic said on Friday night that the U.S. government had ordered the company to suspend access for all foreign nationals to its Fable 5 and Mythos 5 artificial intelligence models, citing national security âconcerns.

> The company revealed the order in a social media post, saying that it âoemust abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.â Foreign nationals who work for Anthropic are also affected by the suspension.

Submission + - 'Dave Eggers doesn't need a smartphone, the internet or your Flock camera' (sfgate.com)

Submitted by destinyland
destinyland writes: Without a pen and paper handy, he was stuck texting the idea to himself. The problem? Eggers doesn't own a smartphone. "It takes 20 minutes to write a sentence," Eggers said... It's a funny predicament for Eggers, given that he's arguably the city's biggest proponent of the written word... Now age 56, Eggers’ latest book is called "Contrapposto "...

On writing days, Eggers bikes to his sailboat docked near the Golden Gate Bridge. He writes using a hefty 1998 Mac that has never been connected to the internet. On the boat, he keeps "banker's hours," working 9 to 5 without any meetings or interruptions except for the occasional wildlife visit. "You're there with the cormorants and the occasional porpoise and sea lions and seals, and when you want to take a break, you walk around and you're in the thick of it, one of the most beautiful spots on Earth," he said. "Especially coming from the Midwest, it never gets old."

Given Eggers' decidedly low-tech existence, it's not surprising that the current state of San Francisco gives him pause, but there's a streak of hope that underlies his concerns. He abhors the growing surveillance technology that's gripping the city, refusing to get into Ubers that use recording devices, but he feels a well-written ballot measure about Flock cameras could potentially save our dwindling privacy. ChatGPT's effects on the art of writing are demoralizing, but he welcomes that teachers are re-embracing pencil and paper, with cursive making a big comeback. The wave of artificial intelligence ads blanketing bus stops imploring companies to stop hiring humans are so over the top, they'd sound cliché if he were to include them in one of his dystopian tech industry novels like "The Circle" or "The Every," but tech philanthropy has helped many of his projects flourish.

Case in point, Art + Water, a new art space scheduled to open next year on Pier 29 funded largely by art world donations... Co-founded with the artist JD Beltran, the space is slated to operate as an old-school apprenticeship system, hosting 10 artists in residence mentoring 20 students, all free of charge... The ultimate goal is to break down the financial barriers that keep students from pursuing art.

Submission + - US Orders Anthropic to Suspend Fable 5 Access (anthropic.com)

Submitted by Anonymous Coward
An anonymous reader writes: Anthropic has announced that the US government issued an emergency export control directive forcing them to abruptly suspend all access to their newly released Fable 5 and Mythos 5 models. Citing national security authorities, the government's order officially targets foreign national access, but Anthropic notes it had to disable the models for all customers globally to ensure full compliance.

The government's concern stems from a verbal report of a potential "jailbreak" method involving code analysis. Anthropic strongly pushes back against the decision, stating that the narrow capability is already common across rival platforms like OpenAI's GPT-5.5. They warned that applying this aggressive standard could effectively halt all frontier model deployments across the AI industry.

Submission + - Autistic kids are being experimented on (theguardian.com)

Submitted by fjo3
fjo3 writes: Across the US, children with autism as young as 18 months old are being given unapproved stem cell treatments at clinics in Florida, Texas and elsewhere, part of a growing market operating beyond the bounds of FDA approval.

The procedure often involves the child being sedated before receiving intravenous doses of millions of stem cells commonly derived from human umbilical cords harvested at birth.

In some cases, the doctors selling the treatments have no scientific expertise in autism or child development. Instead, physicians from unrelated specialties, including plastic surgery and orthopaedics, have entered the booming stem cell sector, billing the procedures as “regenerative medicine” for children, some of whom have severe disabilities

Submission + - Pacific Islanders Appear To Have Most Ancient Human DNA On Earth (studyfinds.com)

Submitted by fjo3
fjo3 writes: People living in Near Oceania, a region spanning New Guinea, surrounding island chains, and the main Solomon Islands, carry more ancient, pre-modern human DNA than any other population on Earth. New research has found that their ancestors inherited genetic material from three distinct Denisovan-like groups, extinct relatives of modern humans, across tens of thousands of years of prehistory. Some of that ancient inheritance appears to influence gene activity today, particularly in immune-related pathways.

Submission + - Anthropic Claims US Government Forced Shutdown of New AI Models (nerds.xyz)

Submitted by BrianFagioli
BrianFagioli writes: Anthropic says the U.S. government ordered it to suspend access to its newest AI models, Fable 5 and Mythos 5, citing national security concerns related to an alleged jailbreak technique. The company says it received the directive Friday evening and was forced to disable access for all users while it works to comply. Anthropic argues the reported jailbreak is narrow rather than universal and claims similar capabilities already exist in competing models, including OpenAI’s GPT-5.5.

The AI company is openly challenging the government’s reasoning, warning that if the same standard were applied across the industry it could effectively halt deployment of future frontier AI models. Anthropic says it supports government oversight of advanced AI systems but argues that any intervention should be transparent, technically justified, and subject to a clear process. The move could mark the first major instance of the U.S. government effectively recalling a frontier AI model after deployment. Readers can find the full story at NERDS.xyz.

Submission + - Data Center Opponents Have Blocked or Delayed Projects Worth Nearly $130 Billion (nbcnews.com)

Submitted by Anonymous Coward
An anonymous reader writes: The first quarter of 2026 produced the most blocked and delayed data center projects on record, according to a new study shared with NBC News. The study — conducted by Data Center Watch, a project of the AI intelligence firm 10a Labs that tracks local data center activity — found that data center opponents blocked or delayed at least 75 projects nationwide worth about $130 billion from January through March, the most in a three-month period since the group began tracking in 2023.

“The quarter reflected a structural shift rather than a cyclical spike: communities have internalized an opposition playbook, legislative sessions introduced formal regulatory uncertainty, and the number of active opposition groups more than doubled to 833 across 49 states,” the authors wrote, noting that the total number and value of data centers blocked or delayed during the first three months of 2026 roughly matched the total for all of 2025.

[...] The report found that legislative pushes for moratoriums on constructing data centers ballooned during the first quarter of 2026, sponsored by lawmakers on both sides of the aisle. The report found such proposals introduced in 14 states from January through March, with Sen. Bernie Sanders, I-Vt., and Rep. Alexandria Ocasio-Cortez, D-N.Y., introducing a federal version. Though none of the proposals has been signed into law, one did reach the desk of Democratic Gov. Janet Mills in Maine. She vetoed it in April.

More than 300 bills were introduced in statehouses across the country just in the first six weeks of 2026, the authors found, saying it marked “a clear shift from incentive-focused policies toward regulatory oversight as the scale of energy demands became clearer.” What’s more, the study found that the number of active grassroots opposition groups across the country more than doubled from 396 at the end of 2025 to 833 by March. The authors found that the states with the most opposition groups through that month were Maryland, Ohio and Texas. “In some cases,” they wrote, “opposition mobilized before any project was officially filed, the mere rumor of a data center was enough to trigger organized resistance.”

Submission + - Arch Linux's AUR Sees More Than 400 Packages Compromised With Malware. (phoronix.com)

Submitted by couchslug
couchslug writes: Michael Larabel reports:

"The Arch Linux User Repository "AUR" was hit by a large-scale malware campaign this week with more than 400 of these user-supplied packages being compromised.

Since yesterday Arch Linux maintainers have been working to reset/delete all of the malicious content and banning affected accounts. Over 400 packages are believed impacted by this latest malware campaign for Arch Linux's AUR. Again, to be completely clear, this just is affecting AUR packages and not the official Arch Linux packages. "

Submission + - Google Sues Chinese-Cybercrime Operation That Used Gemini AI To Send Scam Texts (techcrunch.com)

Submitted by Anonymous Coward
An anonymous reader writes: Google is suing to dismantle the infrastructure behind an alleged massive AI-powered cybercrime operation. On Friday, the tech giant announced a lawsuit against an alleged Chinese cybercrime network called Outsider Enterprise, which Google says uses AI in its campaigns to send scam text messages impersonating Google and other brands to steal passwords and credit card numbers.

Outsider Enterprise has financially scammed “hundreds of thousands of victims” with losses “estimated in the millions.” The group deployed 9,000 fake websites, 1 million fraudulent web domains, and 2.5 million texts sent to Android users in a two-week period, according to Google. “55,000 spam texts were flagged by Android users in just two weeks this past May — that’s more than two text spam complaints a minute,” Google said.

Google said it uses “AI-powered tools to fight AI-powered scams”, which enable the company to detect scams and alert users of suspicious calls and text messages, leading to the interception of more than 10 billion scam messages a month. The company said it has been collaborating with AT&T, T-Mobile, and Verizon to block the scam text messages and said it is coordinating with the FBI, which is taking unspecified law enforcement actions.

Submission + - WAPO sued, reader accuses it of using 'surveillance pricing' to gouge readers (the-independent.com)

Submitted by schwit1
schwit1 writes: Chelsea Bink thought she was buying a subscription. The lawsuit says she was also feeding a pricing machine. From the Independent:

A Washington Post reader has sued the Jeff Bezos-owned newspaper, accusing it of spying on its own subscribers to jack up their subscription prices.

Chelsea Blink’s class action complaint alleges that The Post began "covertly harvesting" data from its subscribers' phones, computers and tablets after the billionaire Amazon founder bought it for $250 million in 2013.

The Post then aggregated and analyzed the "deeply personal information" to "weaponize" it and maximize profits, according to the 28-page lawsuit filed in Superior Court in Washington, D.C.

"The more loyal a reader became, the more data The Post could gather to estimate how much more that person might tolerate paying at renewal," the court filing says. "Rather than rewarding loyalty, The Post’s system converted Subscribers’ engagement into leverage against them. Longtime Subscribers would end up paying more than new customers simply because the company knew more about them."

Blink's lawsuit, first reported by Mediaite, accuses The Post of violating local consumer protection law through its alleged "unfair and deceptive acts."

Submission + - Microsoft Surface firmware left embedded controller unprotected (theregister.com)

Submitted by Dotnaught
Dotnaught writes: For the past 90 days, Microsoft has been quietly patching a firmware flaw in Surface devices that allowed the hardware to be bricked with a single packet, though only for those who have disabled Secure Core and Secure Boot.

And the company's Copilot AI software inadvertently helped identify the faulty firmware. Asked by a security researcher to adjust the backlighting on a Surface laptop, the AI sprayed the embedded controller with data and bricked his device.

Submission + - Usenet is back! (sort of] (newsgrouper.org) 1

Submitted by Anonymous Coward
An anonymous reader writes: Newsgrouper is a free web-based interface for reading and posting to Usenet discussion groups (text only, no binaries). Hosted at newsgrouper.org, it allows users to access Usenet newsgroups through a simple browser interface — no dedicated newsreader software or Usenet provider subscription needed

Key features:

Read and post to Usenet newsgroups via the web
Text-only — no binary (file) groups supported
Guest access available for browsing; account required for posting

It was built as a personal project and shared on Reddit and Hacker News in late 2024/early 2025, with the goal of making Usenet's remaining worthwhile discussion corners (like comp.lang.* groups) more accessible

Submission + - Shutterstock is embracing AI slop and calling it creativity (nerds.xyz)

Submitted by BrianFagioli
BrianFagioli writes: Shutterstock has unveiled what it calls a âoehuman-led, AI-poweredâ creative platform that combines its library of contributor-created content with AI image generation, AI editing, conversational search, prompt enhancement, and automated model selection tools. The company says the goal is to help creators move from idea to finished work faster while maintaining commercial licensing protections and contributor royalty payments.

Critics may see the announcement differently. While Shutterstock repeatedly emphasizes human creativity, much of the platformâ(TM)s future appears centered on AI-generated and AI-modified content. The move highlights a growing tension across the creative industry as companies race to embrace artificial intelligence while creators worry that the internet is becoming increasingly flooded with what many have come to call âoeAI slop.â

Submission + - Euro-Office 1.0 Arrives To Open-Source Infighting (zdnet.com)

Submitted by Anonymous Coward
An anonymous reader writes: If digital sovereignty is important to you, and it certainly is in the European Union (EU), then you'll be pleased to know that EuroOffice, a new open-source browserbased office suite alternative to Microsoft 365 and Google Workspace, has officially reached its first stable release. A coalition of EU-based companies, including Nextcloud, Ionos, and other Euro-Stack participants, is positioning Euro-Office as a cornerstone of European digital sovereignty. However, The Document Foundation (TDF), LibreOffice's steward, accuses the project of reinforcing Microsoft's document lock-in, which TDF argues isn't friendly to open standards.

Setting aside the open-source politics for the moment, here's what Euro-Office brings you. The release went live on June 9. It is, however, not a stand-alone office suite. As the software's backers explain in a FAQ, "Euro-Office is more of an integration component. It merely handles document editing itself. Storage, as well as navigation, permissions, and sharing logic, have to be offered by a platform it is integrated in, like Proton Docs, Nextcloud Hub, or OpenProject." So, while you can install Euro-Office on your own Linux server, you'll need to integrate it yourself. If you're not a Linux expert, however, don't give up hope. Some companies have already released packaged, ready-to-install Euro-Office stacks, including Nextcloud Hub 26 Spring, Ionos' Nextcloud Workspace, and Office.eu. These initial deployments are web-based rather than standalone desktop suites.

The goal, organizers say, is to give European organizations a way to host their office suite on EU infrastructure under EU law, while maintaining an experience familiar to Microsoft Office users. Specifically, Euro-Office is meant to be "a solution for editing documents, spreadsheets, and presentations, developed as a true sovereign community collaboration of over a dozen different organizations."

Submission + - Germany makes landmark decision on Google's AI Overviews (the-decoder.com)

Submitted by Morpeth
Morpeth writes: A German court made clear distinction between Google simply returning search results which point to websites they did not generate/control/own, versus the information provided by Google's AI Overviews, which the court deems as content they are creating, and hence liable for.

"A German court has ruled that Google is directly liable for what its AI search overviews say. Previous case law shielding search engine operators from liability doesn't apply to AI overviews..."

"Google's AI overviews work nothing like traditional search results, the court argues. The AI rewrites and judges results "in its own words and according to its own structure," the ruling says. "

"The court also examined existing rulings from Germany's Federal Court of Justice (BGH), which gave traditional search engines and autocomplete limited liability. The BGH had argued that search engine operators were only liable as indirect infringers because they merely made third-party content findable. A proactive duty to check results would threaten how search engines work.

The Munich court found that this reasoning doesn't apply to AI overviews. A regular search engine just points to outside websites. But AI overviews generate "independent, new, and substantive statements"

Submission + - Facial Recognition Falsely Identifies Florida Man as a Child Abductor (reason.com)

Submitted by fjo3
fjo3 writes: Police arrested a man in Florida for attempted child abduction in a town he had never visited, and the only evidence linking him to the crime was an AI facial recognition hit. Represented by the American Civil Liberties Union (ACLU), he is now suing the officers and agencies who put him through it.

Submission + - HP: Hackers Are Turning Legitimate Remote Access Tools Into Backdoors (nerds.xyz)

Submitted by BrianFagioli
BrianFagioli writes: HP is warning that hackers are increasingly abusing legitimate remote access tools such as LogMeIn and ScreenConnect instead of relying solely on traditional malware. According to the companyâ(TM)s latest Wolf Security Threat Insights Report, attackers have used tax-themed phishing emails, fake software updates, and bogus app downloads to trick users into installing authentic remote access software that ultimately gives cybercriminals persistent control over their PCs. Because the software is legitimate and digitally signed, the activity can blend in with normal IT operations and avoid raising suspicion.

The report also highlights a growing number of attacks involving fake cryptocurrency wallet recovery tools, AI-assisted âoevibe-codedâ malware, and ClickFix campaigns that disguise malware as audio files behind realistic CAPTCHA prompts. HP says email remains the top malware delivery method, accounting for 57 percent of threats observed during the first quarter of 2026. The company argues that modern attackers are increasingly hiding behind trusted software, familiar workflows, and convincing social engineering rather than obviously malicious programs. What do you think about attackers abusing legitimate tools instead of creating their own malware? Is user education enough, or do operating systems need stronger protections against this sort of abuse?

Slashdot Top Deals